Google's own help pages document far lower requirements than the industry repeats. Target CPA can be started with no conversion history at all, and Target ROAS on Search and Shopping asks for 15 conversions in the past 30 days, not 50. The thresholds below are quoted from Google's documentation and ordered from the strategy that needs the least conversion data to the one that needs the most, so you can find the highest rung your account can actually stand on.
Ask how many conversions Smart Bidding needs and you will be told 30 for Target CPA and 50 for Target ROAS, with total confidence and no citation. Both numbers are wrong, or at least neither says what it is claimed to say. We checked every figure below against Google's own help pages rather than against other people repeating them.
How many conversions does Smart Bidding actually need?
Less than the received wisdom claims. Google documents no conversion history requirement for Target CPA and fifteen conversions for Target ROAS on Search.
Ranked by
The conversion data each strategy needs before it can operate, taken from Google's own documentation. Ordered from least to most, so you can start at the top and move down as data accumulates.
Manual CPC and Maximize clicks
First because they need no conversion data at all. Neither one optimises toward conversions, so there is nothing for a shortage of conversions to break.
This is the honest answer for an account with almost no conversion volume, and it is unfashionable enough that most agencies will not offer it. If you are recording four conversions a month, an algorithm cannot learn anything from four data points, and handing it the bidding produces confident-looking decisions built on noise.
The cost is real: you are buying clicks rather than outcomes, and you have to do the work of deciding which keywords, times and devices deserve more of the budget. That is a reasonable trade while the account is too small for anything else, and it is a bad long-term destination.
Maximize conversions, with no target
Second because it needs conversions to exist but Google publishes no minimum volume for it on Search. It spends the budget you give it and reports what that bought.
The absence of a published threshold is not the same as the absence of a requirement. With very few conversions this strategy will still spend the full budget, because spending the budget is what it is designed to do, and the cost per conversion it lands on is whatever the auction gave it rather than a number anyone chose.
It is a reasonable staging post: it teaches the account to record conversions at some scale, which is the input every later strategy needs. Treat it as a way of accumulating data, not as a way of controlling cost.
Target CPA
Third, and lower than almost anyone places it. Google states plainly that you can start with no conversion history. The famous 30 conversions is a recommendation for evaluating performance, not a gate on using it.
This is the single most repeated error in the field. The 30 figure is real and it is on the page, but it describes how much data you want before judging whether the strategy is working. It is not a requirement to switch it on, and Google says so in the same document.
That distinction changes what you do. If you have been staying on manual bidding because you were told you needed 30 conversions a month to qualify, you have been waiting for permission that was never required. What you cannot do is switch to Target CPA and read its performance after a fortnight and eleven conversions, because that is exactly the judgement the 30 figure is about.
“Advertisers can start using Target CPA with no conversion history. For evaluation, we recommend you measure performance for the last 30 days, including at least 30 conversions.” Google Ads Help, About Target CPA bidding, checked 29 July 2026.
Target ROAS on Search and Shopping
Fourth because it carries a published floor, and it is 15 conversions in the past 30 days rather than the 50 usually quoted. It also needs a second kind of data that the others do not.
The volume bar is lower than the industry believes, but the real requirement is the one nobody mentions: Target ROAS optimises toward conversion value, so every conversion has to carry a value, and that value has to be roughly true. An account passing a value of 1 on every lead is giving the algorithm a worse version of Target CPA with extra steps.
If your conversion values are placeholders, or every lead is worth the same because nobody has worked out what a lead is worth, fix that before changing the bid strategy. The value model is the input, and no bidding strategy improves a wrong one.
“Search and Shopping campaigns: At least 15 conversions in the past 30 days at the conversion tracking level.” Google Ads Help, About Target ROAS bidding, checked 29 July 2026.
Value-based bidding on Demand Gen
Last because it carries the highest documented floor on this list by a wide margin, and unlike the others the requirement is specified in both a volume and a recency dimension.
This is the clearest illustration of the general rule. Where Google actually requires volume, it says so precisely, and the number is nothing like the vague thresholds passed around for Search. Everything below this rung has either no published floor or a low one.
If your account is nowhere near this, that is not a failure. It means Demand Gen value bidding is not the tool, and reading the requirement honestly saves you a quarter of watching a campaign fail to leave the learning phase.
“At least 50 Demand Gen conversions with value within the past 35 days, including a minimum of 10 conversions with value within the past 7 days.” Google Ads Help, About Maximize conversion value bidding, checked 29 July 2026.
Why do the numbers everyone quotes not match Google's documentation?
Because the widely repeated thirty and fifty figures come from agency blogs citing each other, not from the help pages they claim to summarise.
The 30 figure has a real origin: it is on Google's Target CPA page, attached to evaluating performance. Somewhere it got restated as an eligibility requirement, that restatement was easier to write than the original, and it propagated. The 50 figure for Target ROAS appears to have no basis in the current documentation at all, which says 15 for Search and Shopping.
This is worth more than a correction about two numbers. It is a working example of how a whole industry can agree on a fact that nobody has checked, and the same thing is true of plenty of other received wisdom in paid search. When a number matters to a decision you are making, spend the four minutes finding it in the platform's own documentation. It is usually one search away and it is surprisingly often different.
One live caveat, from the same pages: Google is renaming these strategies. Both help pages state that from June 2026, "Maximize conversions with a Target CPA" becomes "Target CPA" and "Maximize conversion value with a Target ROAS" becomes "Target ROAS". The behaviour is unchanged; only the labels in the interface move.
What should you do if you are below every threshold?
Fix what you count before you change how you bid. Most accounts below the thresholds have a conversion definition problem rather than a volume problem.
The common situation is an account recording eight conversions a month where the conversion action is a demo request, while a form download, a phone call over 90 seconds and a pricing page visit go uncounted. That account may well have the volume; it is throwing most of it away at the definition stage. Which actions are worth counting, and in what order, is the subject of conversion actions ranked by how well they predict revenue.
If the volume genuinely is not there, the honest answer is that bidding is not your constraint and no bid strategy will become one. Work out what a customer is worth and what that permits you to pay, which the target CPL calculator does from your own numbers, and spend the effort on the offer, the landing page and the keyword set instead.
Does hitting the minimum mean the strategy will work?
No. A documented minimum is where the strategy becomes eligible, not where it becomes reliable, and the gap between those two is usually months.
A threshold is a floor, and floors are where you stand, not where you aim. An account that has just crossed 15 conversions in 30 days can run Target ROAS, and it will also swing hard on small samples, react to a fortnight of unusual traffic as though it were a trend, and take a long time to settle after any change.
The practical test is not the count, it is whether a change produces a signal you can distinguish from noise. That is the same question a conversion test asks, and it has the same uncomfortable answer for most accounts, which the A/B significance calculator will tell you plainly. If your volume cannot evaluate a landing page change, it cannot evaluate a bid strategy change either.
None of this is fixable by choosing a cleverer strategy, which is why the first pass of any account review we do is the measurement rather than the bidding. What gets counted, and whether it is counted correctly, decides the ceiling for everything above it. That is the argument in full on marketing analytics.
Google Ads management
Google Ads management means owning the whole loop: what the account is allowed to bid on, what counts as a conversion, and what the numbers mean once they arrive. We do all three. Most underperforming accounts we see are not badly optimised, they are badly measured, and no amount of bid tuning fixes a conversion action that fires on a thank-you page reload.
Read about Google Ads management